The rise of AI in medicine is both exhilarating and deeply unsettling, especially when you consider its impact on the next generation of doctors. Personally, I think we’re standing at a crossroads where technology could either elevate medical training or inadvertently undermine it. What makes this particularly fascinating is how AI tools like OpenEvidence are reshaping the very foundation of clinical reasoning—a skill that, until now, has been honed through years of struggle, failure, and hands-on experience.
One thing that immediately stands out is the paradox of convenience. Trainees today can bypass the painstaking process of diagnosing a patient by simply querying an AI. On the surface, this seems like a win—faster answers, fewer mistakes, and a more confident appearance on the wards. But if you take a step back and think about it, this convenience comes at a cost. The struggle to diagnose, to grapple with uncertainty, and to learn from mistakes is the crucible of medical training. Without it, we risk creating a generation of doctors who are technically proficient but clinically shallow.
From my perspective, the issue isn’t just about deskilling—it’s about never-skilling. A doctor who’s forgotten how to reason can relearn; a doctor who never learned in the first place is in uncharted territory. What many people don’t realize is that clinical reasoning isn’t just a checklist of facts; it’s an intuitive, almost artistic process shaped by years of exposure to the messy reality of patient care. AI can’t replicate that—at least not yet.
What this really suggests is that we need a radical rethink of how we integrate AI into medical training. It’s not enough to say, ‘Use AI responsibly.’ That’s like telling a pilot to ‘fly safely’ without teaching them how to handle turbulence. We need structural changes—clear guidelines on when and how trainees should use AI. For instance, requiring them to make an unaided diagnosis before consulting the machine could force them to engage their own reasoning first.
A detail that I find especially interesting is the analogy to aviation. Pilots aren’t taught to avoid autopilot; they’re taught to maintain manual flying skills. Medicine needs a similar approach. Trainees should periodically work through cases without AI, not just to test their knowledge but to reveal any gaps in their reasoning. This raises a deeper question: How do we ensure that AI augments human judgment rather than replacing it?
In my opinion, the answer lies in teaching trainees to interrogate AI, not just rely on it. Imagine running drills where they analyze AI-generated assessments for subtle flaws. This wouldn’t just test their knowledge—it would teach them to think critically about the tool itself. What makes this approach compelling is that it doesn’t demonize AI; it positions it as a partner in learning, not a crutch.
If you ask me, the real challenge isn’t the technology itself but how we adapt to it. AI is here to stay, and it has the potential to revolutionize medicine. But patients don’t just need doctors who can operate AI—they need doctors who can think independently of it. That’s the kind of clinician medical training should aim to produce: someone who can stand apart from the machine long enough to know when it’s wrong, incomplete, or right for the wrong reason.
What’s often misunderstood in this debate is that difficulty in training isn’t about hazing—it’s about building resilience and depth. The struggle to diagnose a patient isn’t a rite of passage; it’s a necessary step in developing the kind of judgment that AI can’t replicate. As we integrate AI into medical education, we need to preserve that struggle, not eliminate it.
In the end, the goal isn’t to make medical training harder or easier—it’s to make it smarter. AI should be a tool that sharpens clinical reasoning, not a shortcut that bypasses it. If we get this right, we could create a new generation of doctors who are both technologically savvy and deeply human. But if we don’t, we risk losing something irreplaceable: the art of thinking for ourselves.
